shawgpt-ft / README.md
irenewds's picture
irenewds/shawgpt-ft
c56fb40 verified
|
raw
history blame
2.25 kB
metadata
base_model: TheBloke/Mistral-7B-Instruct-v0.2-GPTQ
library_name: peft
license: apache-2.0
tags:
  - generated_from_trainer
model-index:
  - name: shawgpt-ft
    results: []

shawgpt-ft

This model is a fine-tuned version of TheBloke/Mistral-7B-Instruct-v0.2-GPTQ on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.8528

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 0.0001
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 2
  • num_epochs: 16
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
4.6193 0.9231 3 4.1003
4.3201 1.8462 6 3.8086
3.9755 2.7692 9 3.5351
2.7376 4.0 13 3.1911
3.3753 4.9231 16 2.9601
3.0944 5.8462 19 2.7577
2.8686 6.7692 22 2.5897
2.0042 8.0 26 2.3885
2.4887 8.9231 29 2.2642
2.3262 9.8462 32 2.1536
2.2109 10.7692 35 2.0500
1.5436 12.0 39 1.9406
1.9916 12.9231 42 1.8923
1.9223 13.8462 45 1.8639
1.4561 14.7692 48 1.8528

Framework versions

  • PEFT 0.13.2
  • Transformers 4.44.2
  • Pytorch 2.5.0+cu124
  • Datasets 3.0.2
  • Tokenizers 0.19.1